Semiconductor memory devices are classified into volatile memory devices, which lose stored data at power-off, and non-volatile memory devices, which retain stored data even at power-off.
Volatile memory devices may include a Static Random Access Memory (SRAM), which stores data using latches, and a Dynamic Random Access Memory (DRAM), which stores data using capacitors, according to a data storing manner. SRAM has low integrity and capacity compared to DRAM. But, SRAM is mainly used as cache memory of memory controller because its peripheral circuits are simple in configuration and it operates in high speed.
Semiconductor elements become fine according to advance in semiconductor fabrication technologies. This causes increase in distributions of basic process characteristics of semiconductor elements. For example, in case of SRAM, there are increased distributions of characteristics necessary for designing such as Static Noise Margin (SNM), write margin, sense margin, and the like. Increase in distributions causes difficulty in developing SRAM and decreases stability of memory cells. As a result, the yield may decrease.
Approaches of solving the above-described problems may include a method of improving process distributions. But, it is limited to increase the stability of memory cells and the yield through the method of improving process distributions.